The rDNA Transcription Machinery Is Assembled during Mitosis in Active NORs and Absent in Inactive NORs Pascal Roussel,* Chantal Andr6,* Lucio Comai, I~ and DaniEle Hernandez-Verdun* *Institut Jacques Monod, 75251 Paris Cedex ,50 France;*Laboratoire d'Immunologie Biologique, H6pital Henri Mondor, 01049 Cr6teil, France; §Howard Hughes Medical Institute, Department of Molecular and Cell Biology, University of California ta' Berkeley, Berkeley, California ;4023-02749 and 1lDepartment of Molecular Microbiology, HMR-509, University of Southern California School of Medicine, Los Angeles, California 33009 Abstract. In cycling cells, the rDNAs are expressed interphase cells as well as in mitotic cells. These results from telophase to the end of G2 phase. The early re- clearly indicated that the RNA polymerase I complex sumption of rDNA transcription at telophase raises the remained assembled during mitosis. In addition, RNA D question of the fate of the rDNA transcription machin- polymerase I and the transcription factors varied in the ow n ery during mitosis. At the beginning of mitosis, rDNA same proportions in the positive NORs, suggesting sto- loa d transcription is arrested, and the rDNAs are clustered ichiometric association of these components. The fact ed in specific chromosomal sites, the nucleolar organizer that the rDNA transcription machinery is not equally from h regions (NOR). In human cells, we demonstrate that distributed among NORs most likely reflects the impli- ttp the rDNA transcription machinery, as defined in vitro, cation of the different NORs during the subsequent in- ://rup is colocalized in some NORs and absent from others terphase. Indeed, we demonstrate that only positive ress.o whatever the mitotic phase: RNA polymerase I and the NORs exhibit transcription activity at telophase and rg RNA polymerase I transcription factors, upstream that the level of transcription activity is related to the /jcb/a binding factor and promoter selectivity factor (as veri- amount of rDNA transcription machinery present in rticle fied for TATA-binding protein and TATA-binding the NOR. We propose that assembly of rDNA tran- -pd f/1 protein-associated factor for RNA polymerase I 110), scription machinery preceding mitosis determines ex- 33 /2 were colocalized in the same NORs. The RNA poly- pression of the rDNAs at the beginning of the next cell /23 5 merase I complex was localized using two different an- cycle. Consequently, the association of rDNAs with the /12 6 tibodies recognizing the two largest subunits or only the rDNA transcription machinery defines the "active" 55 8 3 third largest subunit, respectively. These two antibodies NORs and the level of activity at the transition telo- /2 3 5 immunoprecipitated the RNA polymerase I complex in phase/interphase. .p d f b y g u e st o n 1 I N eukaryotic cells during interphase, nucleoli are the (TIF) IB, is a complex composed of the TATA-binding 1 M a sites of rDNA transcription, rRNA processing, and protein (TBP) and three TBP-associated factors (TAF) (7, rch the assembly of ribosomes (19, 50). In human cells, 8, ,11 39, 54). UBF and SL1 most likely represent the mini- 20 2 3 the rDNA transcription machinery, as defined in vitro, is mal set of factors needed in addition to RNA pol I to di- composed of RNA polymerase I (RNA pol 1) 1 in associa- rect rDNA transcription in human cells. Indeed, in addi- tion with the RNA pol I transcription factors upstream tion to UBF and SL1 (designated TIF-IB in mice), two binding factor (UBF) and promoter selectivity factor transcription factors designated TIF-IA and TIF-IC were (SL1) (2, ,3 24). The latter transcription factor or the cor- identified in mouse cells. TIF-IA is a growth-dependent responding murine factor, transcription initiation factor transcription initiation factor. It interacts physically with RNA pol I and converts it into a transcriptionally active RNA pol I (41). TIF-IA was purified from exponentially Address all correspondence to Dani~le Hernandez-Verdun, Institut growing mouse cells and identified in HeLa cells (42). TIF- Jacques Monod, 2 place Jussieu, 75251 Paris Cedex 05, France. Tel. (33) 1 IC, implicated in both transcription initiation and elonga- 44 27 40 38. Fax: (33) 1 44 27 59 94. e-mail: [email protected]. tion of RNA pol I (44), has not yet been identified in hu- 1. Abbreviations used in this paper: BrUTP, bromo-uridine 5'-triphospate; man cells. DAPI, 4',6-diamidino-2-phenylindole; NOR, nucleolus organizer region; However, even if the partners necessary to promote RNA pol I, RNA polymerase I; RT, room temperature; SL1, promoter se- transcription of rDNAs have been recently well character- lectivity factor; TAF, TBP-associated factor; TAF1, TBP-associated factor ized in vitro, this characterization is not sufficient to pre- for RNA polymerase I; TBP, TATA-binding protein; TIF, transcription dict how the rDNA transcription machinery proceeds in initiation factor; UBF, upstream binding factor. © The Rockefeller University Press, 0021-95251961041235112 $2.00 The Journal of Cell Biology, Volume 133, Number 2, April 1996 235-246 235 vivo. In particular, it would be essential to know the fate of ditions to compare the rDNA machinery associated with the rDNA transcription machinery when ribosomal tran- the different types of NORs. In this communication, we scription is repressed. To address these questions, we took demonstrate that the rDNA transcription machinery, as advantage of the physiological conditions provided by the defined in vitro, is colocalized in some NORs and absent cell cycle during which there are alternate phases of ex- from others whatever the mitotic phase. The results show pression and repression of the rDNAs. In cycling cells, the that RNA pol I and SL1 complexes are maintained during rDNAs are expressed from telophase to the end of inter- mitosis. In addition, RNA pol I and the transcription fac- phase and are repressed during mitosis (34). In prophase tors varied in the same proportions in the positive NORs, while rRNA synthesis is down-regulated, nucleoli disap- suggesting stoichiometric association of these components pear. The reformation of nucleoli occurs in telophase at in complexes. specific chromosomal sites designated nucleolar organizer regions (NOR) (25, 33) and requires rDNA transcription Materials and Methods (4, 5, 13). Thus, NORs are the sites where the rDNA tran- scription machinery is activated in telophase and conse- Purified Proteins, Antisera, and Antibodies quently, the sites where the mitotic inactive state should be investigated to understand the transition between mito- The mouse and yeast RNA pol I were provided respectively by .1 Grummt sis and interphase. (German Cancer Research Center, Heidelburg, Germany) and A. NORs are characterized by the presence of rDNA clus- Senenac and M. Riva (CEA, Saclay, France). The mouse anti-TBP mAb (16E8) was obtained from R. Tjian and R. Weinzierl (University of Cali- ters and of several nonhistone proteins. In humans, NORs fornia, Berkeley, CA), and the rabbit polyclonal anti-TBP antibodies pre- are located in the secondary constrictions of the 10 acro- pared against the full-length recombinant human protein were from Santa centric chromosomes (chromosome pairs 13, 14, 15, 21, Cruz Biotechnology (Santa Cruz, CA). The mouse anti-bromodeoxyuri- D and 22) (references ,1 ,21 20, 47). The 400 copies of rDNAs dine mAb was from Boehringer Mannheim France S.A. (Meylan, France). ow n found in humans are distributed in a nonuniform manner The polyclonal rabbit anti-TAF~ll0 (TBP-associated factor for RNA poly loa I 110) antibodies, kindly provided by H. Beckmann (University of Cali- de on these different chromosomes (26). In addition "active" d and "inactive" mitotic NORs were defined by the pres- ftoorniima,m une Berkeley, serum withC A), specificity were described against previousUlBy F (A17) (54). was The described human (38). au- from h ence of a class of proteins designated Ag-NOR proteins. Two other autoimmune sera (A18 and Vll) were characterized in the ttp These proteins were found associated only with the NORs present study. These autoimmune sera contained only autoantibodies ://ru in which the rDNAs were to be efficiently expressed dur- characterized as IgGs and directed against nucleolar proteins. The anti- pre ing interphase (30, 31). These observations indicated that b1o0%d ies SDS-polyacrywleraem ide affinity purified gels using and nucleolar electrotransferred. proteins electrophoresAendt ibodies were in ss.org even during mitosis, in the absence of any transcription eluted by incubation in 100 mM glycine, pH 3, at room temperature (RT) /jcb a(3s4,s ociated 51), there to NORs are differednecpese nding ion n thet heir nonhistone function during proteins in- foboofrd ies, 1 10 M min. Triwsit-hH Cl, The specificity eluted pH 8. fractfioro ns FITC- rabbit and were or Texas mouse neutralized red--conjugateIdg Gs anbdy adding for human secondary 10% IgG(sv,o l/vol) anti- re- /article-pd terphase. Nonhistone proteins associated with NORs dur- spectively, were obtained from Jackson ImmunoResearch Laboratories, f/1 3 ing mitosis have already been identified as components of Inc. (West Grove, PA). Peroxidase-conjugated secondary antibodies with 3/2 the rDNA transcription machinery, such as subunits of specificity for human, rabbit, or mouse IgGs were obtained from Amer- /23 5 RNA pol I (28, 29, 40, 51) and the RNA pol I transcription sham France (Les Ulis, France). /12 6 factor UBF (6, 38, 53), or as being essential for rRNA syn- 55 8 thesis, such as DNA topoisomerase I (16). However, the Preparation of Chromosome-associated Protein 3/2 3 characteristics of the mitotic rDNA transcription machin- Extracts for SDS-PA GE 5.p d eclreya rly are estsiltl ablished poorly understthoaotd . some subunitsF or example, of the even RNA if it pol was I H14 eLa h. Mitcoetlilsc were cells blocked were inh arvested mitosis by by colchicinem echanical treatment shock. The (0.1 ~.g/ml) procedure for f by gu e are localized in NORs, it has never been proven that the used to isolate chromosomes was adapted from Young et al. (52). The st o RNA pol I complex is maintained during mitosis and asso- cells (at least 98% mitotic cells) were centrifuged at 100 g for 10 min, sus- n 1 cnioatt ed yet with been NORs. investigated Similarly, either the during localization mitosis of or SL1 during has pmpioenln.y daemdin e The at 701bcuelflsf er cells were (15 per mM centrifumgle d Tris-HCl, in 75 mM (100 pH KCI, 8, g, 2 and m10 M minicnu)b atEedD TA, and 0.5 at suspendedm M 37°C EGTA, for 15 in 1 March 2 interphase. Moreover, as for RNA pol I, the presence of 80 mM KC1, 20 mM NaCI, 0.2 mM spermine, 0.5 mM spermidine, and 3 02 3 the SL1 complex during mitosis has never been reported. mM dithiothreitol). Cell membranes were then mechanically disrupted by addition of 0.25% (vol/vol) Triton X-100 and passage through a G22 nee- The aim of the present study was to investigate if the dle until most of the chromosomes appeared scattered as assessed by rDNA transcription machinery, as it was defined by in phase microscopy. To discard nonlyzed mitotic ceils and contaminating vitro reconstituted transcription systems, remains associ- nuclei, the chromosome suspension was centrifuged on a 0.25 M sucrose ated with NORs during mitosis. Other questions concern cushion in polyamine buffer at 600 g for 5 min, and the supernatant recen- the respective distribution of RNA pol I and transcription trifuged on a similar sucrose cushion at 34000 g for 30 min. SDS-PAGE sample buffer (27) was added to the pellet containing the chromosomes. factors in the different NORs that could explain the vari- These chromosome extracts were sonicated, boiled for 5 min, and centri- ability of NOR activity during interphase. In addition, the fuged. The proteins in the supernatant corresponding to cytoplasm were stability of this association during mitosis and the partition precipitated by 5 vol of cold acetone and kept at -20°C for 1 h. The pre- between the daughter cells would shed light on the trans- cipitated proteins were collected by centrifugation and solubilized in SDS- PAGE sample buffer. The cytoplasmic extracts were sonicated, boiled for mission of the rDNA transcription machinery through cell 5 min, and centrifuged. cycle. Finally, these results will help to clarify the defini- tion of "active" contrary to "inactive" NORs at least for Preparation of Cell, Nuclear, and Nucleolar Protein the proteins involved in the rDNA transcription. Since the Extracts for SDS-PA GE human NORs differ with respect to the number of rDNA copies and potential activities (9, 10), these are ideal con- All extracts were prepared from exponentially growing HeLa cells. For The Journal of Cell Biology, Volume 133, 1996 236 whole cell extracts, cells were washed in culture medium without serum min. After washing in PBS, cells were permeabilized with acetone at and were lyzed in SDS-PAGE sample buffer. -20°C for 3 min. For nuclear and nucleolar protein extracts, cells were lyzed at 4°C in For immunofluorescence labeling, cells grown as monolayers or chro- TKM buffer (10 mM Tris-HCl, pH 7.4, 10 mM KC1 and 3 mM MgCI2). Ly- mosome spreads were incubated with sera at RT for 45 min, and the anti- sis was stopped when the nuclei appeared free of cytoplasmic components bodies were revealed by Texas red-conjugated goat anti-human and/or as assessed by phase microscopy. The nuclear and nucleolar proteins were FfTC-conjugated goat anti-rabbit antibodies. DNA was visualized with prepared as previously described (37). The nucleoli were isolated by soni- 4',6-diamidino-2-phenylindole (DAPI). All preparations were mounted cation of nuclei suspended in TKM buffer and purified by centrifugation with an antifading solution (Citifluor, Canterbury, UK). on a 0.88 M sucrose cushion in TKM buffer at 1,300 g for 15 min. All steps Fluorescent microscopy was performed using a microscope (DMRB; E. were performed at 4°C, and all of the solutions contained an antiprotease Leitz, Inc., Rockleigh, N J). Images were photographed using a microscope cocktail: 1 mM PMSF, 10 ~g/ml aprotinin, 1 ixg/ml pepstatin, and 1 p.g/ml camera system (DMRD; E. Leitz, Inc.). The superimposition of images leupeptin. The nucleolar proteins were solubilized in SDS-PAGE sample was obtained by scanning micrographs using Ofoto 1.0.1 on Onescanner buffer. (Apple Computer, Inc., Cupertino, CA). Images were then assembled and printed using Canvas (Deneba Systems Inc., Miami, FL) and Adobe Pho- Immunoblotting toshop (Adobe Systems Inc., Mountian View, CA). All computer manipu- lations were done with a Macintosh Quadra 700 (Apple Computer Inc.). The proteins were separated by 10% SDS-PAGE using a Protean II cell Images were printed directly from the computer on a printer (ColorEase (Bio-Rad Laboratories, Richmond, CA). Size standards from 200 to 14 PS Printer; Eastman-Kodak Co., Rochester, NY). kD (Bio-Rad Laboratories) were included in each gel. The polypeptides were electrotransferred to reinforced cellulose nitrate membranes (BA-S Assay of RNA Polymerase Activity In Situ 83; Schleicher & SchueU, Dassel, Germany), that were then cut into strips. The strips were blocked by incubation for 1 h in PBS containing 5% (wt/ The assay was performed on HeLa cells grown as monolayers essentially vol) dried milk and 0.05% (vol/vol) Tween-20 and incubated with the sera as described (32). Cells were incubated in the assay solution (100 mM for 2 h in the same buffer. They were then washed three times with PBS Tris-HC1, pH 7.9, 12 mM 2-mercaptoethanol, 150 mM sucrose, 0.6 mM containing 5% (wt/vol) dried milk and 0.5% (vol/vol) Tween-20 and incu- ATP, CTP, and GTP, 0.12 mM bromo--uridine 5'-triphosphate (BrUTP), D bated for 1 h in the presence of HRP-labeled second antibodies. After and 12 mM MgC12) at 37°C for 15 min. The reaction was terminated by ow several rounds of washing, the HRP activity was detected using the en- rinsing the slides in PBS and postfixing with 2% (wt/vol) formaldehyde in nlo hanced chemiluminescence kit (Amersham France) and recorded on x-ray PBS at RT for 15 min followed by permeabilization with acetone at -20°C ade film (Fuji Photo Film Co., Ltd., Tokyo, Japan). f(o49r) 3 by rain. immunofluorescBernUceT P incorporation labeling was then using detected a mouse as anptrie-vbiroousmloyd eoxyuri- described d from dine mAb revealed by FITC-conjugated goat anti-mouse antibodies. In h Immunoprecipitation HeLa cells were metabolically labeled with 35Smethionine (Tran35S - sveoamleed experimenbtys , Texas red-conjugated UBF was simultaneously goat anti-human detected antibodies. using serum A17 re- ttp://rupre label; ICN Biomedicals, Orsay, France). Interphase cells were scraped off noC focal Laser Scanning ypocsorciM ss.o in ice-cold PBS after eliminating mitotic cells by mechanical shock. Mi- rg thtEwooetatDriireccT v eAs,t ed ceclellsl s, sonicated 1 webrye mMo btained mechanical eixotdroaacactneedtd a mide, by centrifuged colchicine shocink 50 mMa nda nd at treatmweTarnsitsh100,000 - eHdC1 %l , (vol/vol) in g (0.1 pHP BS. for 8, p.g/Tml)r it1 on The 500 h. The imnMt erphaosvXe-e1 0r0.n ight, supNeaCrI,n atants Extracts o5 r were mM mi- C(IAaonBncnn.idf, o o-cRaaid argona Garden helium-neon laisoLenar b oratories), City, laser scanning NY) adijouns ted elqasueirp ped mimcoruosnctopeyd to adjusted 488 with nmo n wast oa wa as ×60 54m3 icroscope perfoursemde d objnecmti ve for for the Texas( Optipuhso(itpn lga n fluorescein red. apoa;n TMhReCII -; 600 NA sNiignkaol,n emit- 1.4). /jcb/article-pdf/133 were adjusted to 50 mM Tris-HCl, pH 8, 2 mg/ml BSA, 150 mM NaCl, 5 ted light was separated by a dichroic mirror (DR565; Bio-Rad Laborato- /2/2 mM EDTA, 1 mM iodoacetamide, and 0.2% (vol/vol) Triton X-100 and 3 were Before used imimnu noprecipitationt,h e immunoprecipitation antibodies assays. were cross-linked to protein A riine s), front and of a t(h54e0D F30; photomultiplier Bio-Rad Labcoorllaetctoirnig es) the long fluorescein pass filter emission. was placed For 5/1265 each optical section, double fluorescent images were acquired in two 5 co(Pviaelrecnet ly Chemical linked Co.t,o agarose Rockford, using IL). the Protein ImmunoPure A-antibody IgG cOomrpileenxteast ion were Kit psaysssetse: m wafs luorescein adjusted first, to allow Texas a fireled d second.d epth of The ~0.5 pinhole ixm, correspondionfg the confocal to 83/235 incubated with extracts at 4°C for 2 h. Immunoprecipitates were washed .p tEwiDceT A, in 1 50 mM mM iodoacetamide, Tris-HCl, pH and 8, 2 1% mg/ml (vol/vol) BSA, Triton 150 mM X-100, NaCt, once 5 in mM 50 tghree en increment and red sbiegntawles en was two collected adjacent for sections. each specimen A focal searnide s then including processed the df by g m6.8. aMnd 1E% Tarcish-H C1, (vol/vol) immunoprecipitation pH Triton 8, 500 X-100, mM NaC1, and reaction finall5 y mM was twicEe DTA, carried in 100 1 mM out mM iodoacetaumsiidneg, Tr is-HC1, extracts pH teloaa tcthe r produce of case, the steresoi-ngpolapeit risc al composite section with or a imagesti lt extended of -0.6/(e+x0t.e6n ded focus imawgeerse focus). generated, were For mecorlgoecdal.i zation, combining In the uest on 11 a high spatial resolution with the observation of the colocalization in the M fmruonme 500,000 sera (A18 cells. and ImmunopreVcilpilta)t es and human nonautoiombmtuainnee d using serhuum man (C) autoim- were cell volume. arch analyzed by 10% SDS-PAGE followed by autoradiography. 20 2 Results 3 Cell Culture and Immunofluorescence Labeling Antibodies against the rDNA Transcription Machinery HeLa cells were cultured in Eagle's minimum essential medium (Flow Laboratories, Paris, France) supplemented with 10% (vol/vol) FCS. Cells The characterization of sera directed against the transcrip- were seeded three times a week, and extracts were prepared 24 h after tion machinery was carried out using nucleolar extracts seeding from exponentially growing cells. prepared from actively growing HeLa cells and partially For immunolocalization, HeLa cells were grown as monolayers on glass slides. The glass slides were previously treated with 5% (wt/vol) KOH in purified RNA pol I complex. The human autoimmune se- methanol for 12 h, rinsed in water, treated with 0.01 N HCI for 1 h, washed rum A18 recognized two polypeptides in HeLa cell nucle- in distilled water and ethanol, and sterilized. Cell monolayers were rinsed olar extracts (Fig. 1 A). The molecular masses of these two in PBS and fixed with 80% ethanol at -20°C for 8 min or with 2% (wt/vol) polypeptides were 200 and 120 kD. Using partially puri- formaldehyde in PBS at RT for 20 min followed by washing in PBS and fied mouse RNA pol I complex (43, 44), these two permeabilization with acetone at -20°C for 3 rain. HeLa chromosome spreading for immunolocalization was carried out polypeptides were identified as the two largest subunits of as follows: HeLa cells were blocked in mitosis by incubation in 0.1 ~g/ml RNA polymerase I (Fig. 1 B). A18 also recognized the two of colcemid for 2 h. Mitotic cells harvested by mechanical shock were cen- largest subunits of the purified yeast RNA pol I (data not trifuged at 100 g for 10 min, suspended at 701 cells per ml in FCS diluted shown). Similarly, serum Vll that recognized a 60-kD 1:5, and incubated at RT for 25 rain. Cells were then cytocentrifuged on glass slides and fixed in 2% (wt/vol) formaldehyde in PBS at RT for 20 polypeptide in HeLa cell nucleolar extracts (Fig. 1 A) rec- Roussel et al. rDNA Transcription Machinery during Mitosis 237 ognized the 62-kD polypeptide corresponding to the third polypeptide could be a proteolytic product of the 60-kD largest subunit of the partially purified mouse RNA pol I subunit. However, the most striking point is the similarity complex (Fig. 1 B). Serum A17, previously characterized of the complexes immunoprecipitated. This similarity ar- as containing anti-UBF antibodies (38) and anti-TBP anti- gues in favor of the fact that both sera are directed against bodies, were tested on these nucleolar extracts (Fig. 1 A). antigens of the same complex that are most likely different UBF, appearing as a doublet of 97-94 kD, and TBP were subunits of the RNA pol I complex. present in the nucleolar extracts. Except for TBP mainly found in nuclear extracts, UBF and the antigens revealed Association of the rDNA Transcription Machinery with by A18 and Vll are concentrated in nucleolar extracts as Chromosomesd uring Mitosis illustrated for V11 (Fig. 2). Immunoblotting seemed to indicate that serum A18 rec- The antibodies were probed on chromosome-associated ognized the two largest subunits of RNA pol I, and serum protein extracts and cytoplasmic extracts prepared from Vll recognized the third largest subunit of RNA pol I. To the same number of mitotic cells. The results obtained confirm the specificity of these sera, immunoprecipitations with serum Vll, anti-UBF, and anti-TBP antibodies are were carried out anticipating that by using antibodies di- presented (Fig. 3). The 60-kD polypeptide revealed by rected against one or two RNA pol I subunits, it would be Vll (corresponding to the 60-kD subunit of the RNA pol possible to immunoprecipitate the entire RNA pol I com- I) was found only in the chromosome-associated protein plex and a similar complex with both sera. At least with re- extracts (Fig. 3, lane a), but not in the cytoplasmic extracts spect to the major bands immunoprecipitated using both (Fig. 3, lane b). Likewise, UBF was observed in the chro- the A18 and Vll sera, the molecular masses correlated mosome-associated protein extracts (Fig. 3, lane c). A very well with those of subunits of the RNA pol I complex typi- weak amount of UBF, only detected when the film was D o cal of mammalian cells (43). The major bands immunopre- overexposed, was present in the cytoplasmic extracts (not w n cipitated from interphase cell extracts using A18 and Vll visible in lane d), most likely due to release during chro- loa d e were 190-, 120-, 52-, 43-, 29-, and 17-kD polypeptides (Fig. mosome isolation. Similarly TBP (Fig. 3, lane e) was d 1 C). Minor bands were observed of 185, 170, 155, and 135 present in the chromosome-associated protein extracts. from h kD, corresponding most likely to proteolytic products of The presence of one subunit of the RNA pol I that asso- ttp the 190-kD polypeptide. These cleavages were also ob- ciated with chromosomes during mitosis did not prove the ://ru p served in enriched mouse RNA pol I extracts (data not association of all the subunits of the complex. Therefore, re shown), in favor of the instability of the highest RNA pol I immunoprecipitations were carried out to investigate the ss.o rg subunit. Serum Vll recognized a 60-kD polypeptide in hu- fate of the RNA pol I complex during mitosis. In mitotic /jcb ma an complex, nucleolar as did extracts A18 in (Fig. which 1 A) no and 60 kD immunoprecipitated was found (Fig. 1 ce(lFlig . extracts, 4 b) and similar Vll (Fig. results 4 c). were The obtaisnaemde major using bands sera were A18 /article-p d C). As the 52-kD polypeptide immunoprecipitated with observed as in interphase but with two additional bands of f/1 3 both sera is more abundant with Vll than with A18, this 75 and 41 kD. The fact that similar patterns were obtained 3/2 /2 3 5 /1 erugiF .1 Identification of 265 5 antigens by immunoblotting 83 and immunoprecipitation. /23 5 (A) HeLa nucleolar protein .pd belxottrsa cts by retvheea led sera A18 on Western (lane a, f by gu e revealing two polypeptides st o n of 002 and 021 kD), Vll 1 1 (lane b, revealing a 60-kD M a polypeptide), anti-UBF anti- rch bodies (lane c), and by anti- 20 2 TBP antibodies (lane d). 3 Numbers indicate the molec- ular masses in kD of proteins revealed with the sera A18 and Vll. (B) Partially puri- fied mouse RNA pol I re- vealed on Western blots by A18 (lane a) and Vll (lane b), recognizing the two larg- est subunits and the third largest subunit of mouse RNA pol I, respectively. (C) Immunoprecipitation assays carried out using control se- rum (C; lane a) and anti- RNA pol I antibodies (A18 and Vll; lanes b and c) on in- terphase HeLa extracts. cell ehT lanruoJ of lleC ,ygoloiB .331 emuloV 6991 832 erugiF .2 Localization of antigens dur- ing interphase by immunoblotting. The same amount of nucleolar (lane a), nu- clear (lane b), and cell protein extracts (lane c) revealed on Western blots with Vll. The position of the 60-kD erugiF .4 Immunoprecipita- polypeptide indicated. is tion assays carried out using control serum (C; lane a) and anti-RNA pol I antibodies using mitotic cell extracts and interphase cell extracts (A18 and Vll; lanes b and c) on mitotic HeLa extracts. cell proves that the RNA pol I complex present during inter- Proteins specifically immu- phase is also present during mitosis. We conclude that the noprecipitated are indicated subunits composing the RNA pol I complex remain associ- by their molecular masses in ated during mitosis. kD. The asterisks point to Do w specifically immunoprecipi- n Localization of the rDNA Transcription Machinery tated bands only observed load e during Mitosis with mitotic extracts. cell d fro m In the chromosome-associated protein extracts of HeLa h mitotic cells, we observed the presence of the RNA pol I As illustrated for early metaphase (Fig. 5 b), labeling ttp://ru subunits, TBP, and UBF by immunoblotting. This is com- p corresponding to TAFIll0 appeared as discrete spots of re patible with the hypothesis that the rDNA transcription variable sizes in association with chromosomes (compare ss.o mTmdiaoetcf ohitniienc ery test if thicsel ls they remahiynpost hesist,h e are majcoolro calized. associated we components proceeded We to firstr DNAs of to detected this immunodetecdtur ing machinery each com- mitosis. toin 5F,ti gh. e c-e) same5, a showed and HeLa b). a mitTothiec similar examination cells localization (illustrated of oUfB F TAFIflorl 0 and telophase, TAFIll0 and UBF Fig. in rg/jcb/article-pd ponent individually. As expected by the results of immu- and similar variations in the size of the spots. Indeed, simi- f/13 noblotting, RNA pol I, UBF, and TBP remained associ- lsamra ller patterns were spots revealed obtained with ainn ti-TAFill0 both cases: antiboditehse larger (Fig. and 3/2/23 ated with the condensed chromosomes at all phases of the 5 d) corresponded respectively to the larger and smaller 5/12 mitosis (data not shown). We also investigated the possi- spots revealed with anti-UBF antibodies (Fig. 5 e). Similar 655 ble colocalization of the transcription factors SL1 and 83 results were obtained whatever the mitotic stage examined /2 UBF that, in addition to RNA pol I, are necessary and suf- 35 (data not shown). .p ficient to promote rDNA transcription in human cells. d The localization of TBP during mitosis in HeLa cells is f b Since SL1 is composed of TBP, TAFIll0, TAFI63, and illustrated for anaphase (Fig. 5 g) and for metaphase (Fig. y gu cTAaFrIr4i8e,d out the using localization anti-TBP of and SL1 anti-TAFi1i1n 0 mitotic HeLa antibodies. cells was 5 j). In both anaphase and metaphase, as was the case at all est on stages of mitosis (data not shown), part of TBP appeared 1 1 in discrete spots in association with chromosomes. As ex- M a pected, part of TBP was not associated with chromosomes rch erugiF .3 Localization of an- but was dispersed in the cytoplasm (Fig. 5 g). The localiza- 202 tigens during mitosis by im- 3 tion of both TBP and UBF in the same mitotic cells (Fig. 5, munoblotting. Comparable f-h, and i-k) showed that at all mitotic stages, the part of amounts of chromosome (lanes a, c and e) or cytoplas- TBP appearing in spots in association with chromosomes mic extracts (lanes b and d) was localized in the same position as UBF. As reported for revealed on Western blots TAFIll0, the spots varied in size, but these variations with Vll (lanes a and b), were similar for TBP and for UBF (compare Fig. 5, g and anti-UBF antibodies (lanes c h, and j and k). The similar localization of TAFIll0 and and d), and anti-TBP anti- TBP indicates that the transcription factor SL1 is most bodies (lane e). The 60-kD probably maintained during mitosis and localized in NORs polypeptide revealed by Vll during mitosis as is UBF (38). The fact that the size of the (lane a) and UBF (lane c) si spots varied in a correlated manner for UBF, TAFIll0, present in chromosome ex- and TBP indicates that the amounts of UBF, TAFIll0, tracts and absent from cyto- plasmic extracts (lanes b and and TBP vary in the same proportions between different d). TBP (lane e) is partially positive NORs. present in chromosome ex- To validate these observations, the colocalization of tracts. these factors in the volume of mitotic cells was established Roussel et al. rDNA Transcription Machinery during Mitosis 239 by confocal microscopy. The fluorescent patterns ob- served in the same mitotic HeLa cells for RNA pol I (us- ing Vll and A18) and TBP (Fig. 6) or for UBF and TBP (data not shown) were identical from prophase to telo- phase. The same number of spots, and spots varying in size in the same proportions, were observed for both labelings in the two cases using extended focus images obtained af- ter three-dimensional reconstruction of the serial optical sections (Fig. 6; illustrated for anaphase, metaphase, and interphase). The superimposition of RNA pol I and TBP labelings recorded in the same optical section indicated strict colocalization of RNA pol I and TBP in mitosis as well as in the nucleolus during interphase (Fig. 6, C, C', and C", and D, D', and D"). In extended focus images six chromosomes were scored positive for RNA pol I colocal- ized with TBP (Fig. 6, A, A', and A") and also for UBF colocalized with TBP (data not shown). Because UBF is known to remain associated with only some NORs during mitosis (6, 38, 53), these results indi- cated that UBF, RNA pol I, and SL1 (verified for TBP and TAFIll0) were associated with the same NORs dur- D o ing mitosis in HeLa cells. wn lo a d e Different Types of NORs d fro m In human cells there are 10 NOR-bearing chromosomes, h but the rDNA transcription machinery remains associated ttp://ru in variable amounts with only six of them. We were inter- p re ested in identifying the positive NOR-bearing chromo- ss.o sgiovmeens cell to detliene.r mine if they constitute a characteristic in a rg/jcb/a Chromosome spreads prepared from HeLa cells blocked rticle in mitosis made it possible to recognize acrocentric chro- -pd mosomes (Fig. 7) that were previously proved to be the f/13 3 NOR-bearing chromosomes in human cells. Immunolabel- /2/2 3 ings of UBF and RNA pol I (Fig. 7) obtained on HeLa 5/1 2 chromosome spreads were similar. In both cases, six spots 65 5 or double-spots were visible. These proteins were not 83 /2 equally distributed between NORs. Some NOR-bearing 35 .p chromosomes scored negative, and the intensity of the la- d f b beling of the six positive NORs differed. Interestingly, y g u UsixB F different and RNA positive pol I NORvasr.i ed in the same proportions in the est on 1 In this HeLa cell line, the labeling pattern was the same 1 M in the different chromosome spreads observed using anti- arch bodies directed against UBF or the RNA pol I subunits 20 2 (Fig. 7). There were always six positive NOR-bearing 3 chromosomes, three small acrocentric chromosomes (chro- These images are reminiscent of background often associated us- ing rabbit antibodies. (f-h) Anaphase HeLa cell labeled with anti-TBP (g) and anti-UBF antibodies (h) and DNA stained by erugiF .5 Localization of UBF and TAFIll0, and of UBF and DAPI (]). TBP labeling (g) appears as discrete spots in associa- TBP, in the same mitotic cells. (a and b) Early metaphase HeLa tion with chromosomes (compare f and g) and as labeling dis- cell labeled with anti-TAFill0 antibodies. The labeling appears persed in the cytoplasm. The discrete spots revealed by anti-TBP in discrete spots (b) in association with chromosomes revealed by coincide exactly with the spots revealed by anti-UBF antibodies DAPI (a). (c-e) Telophase HeLa cell labeled both with anti- .)h( (i-k) Metaphase HeLa cell labeled with anti-TBP (/'a)n d TAFIll0 (d) and anti-UBF antibodies (e) and DNA stained by with anti-UBF antibodies (k) and DNA stained by DAPI (i). A DAPI (c). The colocalization of UBF and TAFIll0 si shown by part of the TBP labeling appears as discrete spots (j) in associa- identical labeling patterns obtained with both antibodies. The tion with chromosomes (i). The spots revealed with anti-TBP an- weak cytoplasmic labeling observed with anti-TAFill0 antibod- tibodies (j) coincide exactly with the spots revealed with anti- ies is homogeneous and visible in mitotic and interphase cells. UBF antibodies (k). Bar, 01 .m~p ehT lanruoJ of lleC ,ygoloiB emuloV .331 6991 042 mosomes 21 or 22), and three large acrocentric chromo- Discussion somes (chromosomes 13, 14, or 15). Thus, for a pair of ho- In vitro studies made it possible to define the rDNA tran- mologous NOR-bearing chromosomes, the first chromosome scription machinery necessary to obtain accurate in vitro could be positive, whereas the second one could be nega- rDNA transcription. The rDNA transcription machinery tive. Moreover, when the labeling of a positive chromo- is composed in humans of the RNA pol I complex, the some appeared as a double-spot (i.e., when the two spots transcription factor UBF, and the transcription factor SL1, corresponding to the two chromatids were not coalesced), composed of TBP and three TAFIs: TAF1110, TAFt63, both spots composing the double-spot were of equal inten- and TAFI48 (2, 3, 7, 24). It was recently proven that the sity. Therefore, the amount of protein associated with three TAF~s together with TBP are necessary and suffi- both chromatids of the same chromosome seemed to be cient to reconstitute a transcriptionally active SL1 com- equivalent. This explained the equal partition of UBF and plex (54). In vivo, expression of the rDNAs is dependent RNA pol I and consequently of the rDNA transcription on cell cycle and cell growth and takes place in the nucle- machinery during anaphase when the separation of chro- oli. During mitosis, rDNA transcription is arrested, but the matids occurs. early resumption of its transcription at telophase raises the Keeping in mind that the different components of the question of the fate of the rDNA transcription machinery rDNA transcription machinery are colocalized during mi- during mitosis. As the NORs are the sites where rDNA tosis, we can conclude that RNA pol I, UBF, and also SL1 transcription is activated in telophase (4, 5, 13, 25, 33), we are associated with the same NORs in a given cell and al- focused our attention on the proteins associated with the ways with the same NORs in this HeLa cell line, and that NORs during mitosis to investigate the transitions be- they vary in the same proportions in the different positive tween mitosis and interphase. NORs. D o w n lo gnitegraT of the rDNA noitpircsnarT Machinery ad The noitpircsnarT Activity of evitisoP NORs e d gnirud sisotiM fro BNeOcRasus e and the requires reformation rDNA of transncurcilepotliio n, occurs we in could telophase predict at For accurate rDNA transcription, it seems reasonable to m http that NORs are the sites where the rDNA transcription postulate that the same partners interact in vivo as in vitro. ://rup vmaancthaignee ry of is the activated. fact that To RNA verify pol this activities point, mawye took be pre- ad- vIinv o this transcription case, the minimal is active transcriptioRn NA pol I mcaocmhpilneexr y and for the in ress.org served in situ in cells (32). During interphase the RNA pol specific RNA pol I transcription factors UBF and SL1. /jcb I activity (insensitive to a-amanitin) occurring in nucleoli mitosis RNA in pol different I has already mammalian been localizecedl ls (14, into 17, 18, NORs 23, 40, during 51). /article is particularly well detected in the conditions used (32). At -p the transition interphase/mitosis, the RNA pol I activity The mammalian subunits composing the RNA pol I com- df/1 3 plex are not as well characterized as the yeast subunits 3 stopped rapidly before the destructuring of nucleoli (data /2 not shown). During mitosis, no RNA pol activity was ob- (45), but recent progress indicated a similar number of /235 served as illustrated for prometaphase HeLa cells (Fig. 8, subunits in mouse and yeast RNA pol I complex (46). In /126 C and C'), except for during telophase (Fig. 8, A, A', A", the present study, we present evidence for the presence of 558 B, B', C, and C'). During telophase RNA pol I activity was the three largest RNA pol I subunits in the NORs by im- 3/23 derepressed as evidenced by the incorporation of BrUTP munolocalization, and by immunoprecipitation we demon- 5.pd paaprpeea ring Fig. 8, in A', spots A", in A, association B', B, C', with and chromC)o. somes The fluorescent (com- splterxaets e that (see mitotibce low forc ell the extracts discussion contain of thReN A mitotic pol I RNAco m- f by gue patterns observed in the same telophase HeLa cells for pol I complex). st o n BrUTP incorporation and for UBF (Fig. 8, B', B", C', and RNA pol I transcription factors were also associated 11 with the mitotic chromosomes as already reported for the M C") showed that both labelings were superimposable. It is a noteworthy that the unusual background obtained using UBF (6, 35, 38, 53). In addition, we proved that SL1 (as rch 2 anti-UBF antibodies was most likely due to the procedure verified for TBP and TAFIll0) is most likely maintained 023 during mitosis and also remains associated with NORs. used to detect RNA pol activity. The localization of UBF, Moreover, all of the rDNA transcription machinery as de- and consequently of the rDNA transcription machinery, fined in vitro is colocalized at NORs during mitosis: RNA and the detection of RNA pol I activity made it possible to pol I, UBF, and SL1 are associated with the same NORs observe that the rRNA transcription machinery was dere- whatever the mitotic stage investigated. We failed to ob- pressed at telophase at NORs. Because up to six spots cor- tain labeling with TAFI48 antibodies in mitosis or in inter- responding to sites of RNA pol I activity could be ob- phase. Because we cannot suspect the absence of TAFI48 served in telophase HeLa cells (Fig. 8, A' and A"), the in active rDNA transcription machinery in interphase, this rDNA transcription machinery associated with the six result is most likely due to inaccessibility of the epitopes or NORs seemed to be derepressed simultaneously. More- modifications induced by fixing the cells. over, since the intensity of the labelings corresponding to the sites of BrUTP incorporation varied in the same pro- The rDNA noitpircsnarT Machinery Is dezilacoloC in portion as the size of the spots corresponding to UBF Active NORs and Absent from Inactive NORs (compare Fig. 8, B' and B", and C' and C"), the level of transcription activity seemed to be directly related to the The active NORs were previously defined as NORs in amount of rDNA transcription machinery present in the which potentially active rDNAs are located. This implies NOR. that these genes can be transcribed during interphase as Roussel e! a. rDNA Transcription Machinery during Mitosis 241 D o w n lo a d e d fro m h ttp ://ru p re ss.o rg /jcb /a rticle -p d f/1 3 3 /2 /2 3 5 /1 2 6 5 5 8 3 /2 3 5 .p d f b y g u e st o n 1 1 M a rch 2 0 2 3 The Journal of Cell Biology. Volume 133, 1996 242 D o w n lo a d e d fro m h ttp ://ru p re ss.o rg /jcb /a rticle -p d f/1 3 3 /2 /2 3 5 /1 2 6 5 5 8 3 /2 3 5 .p d f b y g u e Figure .7 Localization of UBF and RNA pol I by immunofluorescence on HeLa chromosome spreads. HeLa chromosome spreads st o stained by DAPI and labeling obtained with anti-UBF antibodies (A and B) or with Vll (C) were superimposed. The overall picture n 1 1 (A) shows that the labeling pattern is the same when comparing one chromosome spread to another. The comparison between labeling M a obtained with anti-UBF (B) and anti-RNA pol I antibodies (C) demonstrates that the same six acrocentric chromosomes (NOR-bearing rch chromosomes in humans) are labeled. Both labelings vary in the same proportions in the six positive chromosomes. The acrocentric 20 2 chromosomes devoid of either labeling are indicated by arrowheads (B and C). Bar, 5 .mxt 3 opposed to the other rDNA copies that cannot be tran- NOR proteins (30, 31). The Ag-NOR proteins are easily scribed. As demonstrated by experiments using interspe- visualized in situ by cytological silver staining (15, 22). We cific somatic cell hybrids, NORs bearing potentially active previously reported that in mitotic HeLa cells, UBF is only rDNAs are characterized by their association with Ag- localized in active NORs where Ag-NOR proteins are also Figure6. (cid:127)ac(cid:127)fn(cid:127)C r(cid:127)sa(cid:127) scanning yp(cid:127)cs(cid:127)rcim dem(cid:127)nstrating n(cid:127)itazi(cid:127)ac(cid:127)(cid:127)(cid:127)c f(cid:127) RNA (cid:127)(cid:127)p (cid:127) and TBP during mit(cid:127)sis. (A(cid:127)A (cid:127)' and A '') The anaphase HeLa cell was labeled with Vll and anti-TBP antibodies. (A) Extended focus of labeling obtained with Vll. (A') Extended focus of labeling obtained with anti-TBP antibodies. (A") Merged image of both labelings. Yellow color indicates colocalization of RNA pol I and TBP present in six spots in association with chromosomes of each daughter cell. (B-D) Metaphase and interphase HeLa cells were labeled with Vll and anti-TBP antibodies. (B) Three-dimensional stereoscopic projections of optical sections ofb oth label- ings showing the colocalization of RNA pol I and TBP in the cell volume. C, C', and C", and D, D', and D" show the labeling obtained with Vll (C and D) and anti-TBP antibodies (C' and D') in two successive optical sections. (C") corresponds to the merged image of both labelings (shown in C and C') recorded in the first optical section, and D" corresponds to the merged image of both labelings (shown in D and D') recorded in the second optical section. Yellow color indicates strict colocalization of both labelings. Bar, 10 .mxt lessuoR te .la ANDr noitpircsnarT yrenihcaM gnirud sisotiM 342 erugiF .8 Detection of RNA pol I activity in mitotic HeLa ceils. (A, A', A") The telo- phase HeLa was cell processed to detect the sites of BrUTP incorporation. (A) DNA stain- ing. (A' and A") Two differ- ently focused images to allow the observation of the six sites of BrUTP incorporation in as- sociation with the two groups of chromosomes revealed by DAPI (A). (B, B', B") The te- lophase HeLa cell was pro- D o w cessed to detect simulta- n lo neously BrUTP incorporation ad e (anBd ') UBF. (B) BrUTP DNA incorporation. staining. d from (B") UBF labeling. (C, C', C") http Telophase and prometaphase ://ru HeLa cells were processed to pre detect simultaneously BrUTP ss.o DinNcoArp oration staining allowing recog- and UBF. (C) rg/jcb/a nition of the telophase (t) and rticle the prometaphase cell (p). (C') -p d BrUTP incorporation. (C") f/1 3 UBF labeling. The arrows 3/2 point to the labeling observed /23 5 for UBF (C") and the absence /1 2 6 of BrUTP incorporation (C') 5 5 8 in the prometaphase cell. Bar, 3 /2 01 .m~p 35 .p d f b y g u e st o n localized (38). These results are corroborated by the iden- only with the active NORs during mitosis (i.e., six active 1 1 tification of the largest subunit of RNA pol I and UBF as NORs in the HeLa cell line used). In mitotic cells and Ma Ag-NOR proteins (36). Considering the published results chromosome spreads, the number of positive chromo- rch 2 0 (9, 10, 30, 31) and the results obtained in this study con- somes are always the same in a given cell line. This charac- 2 3 cerning the colocalization of RNA pol I, UBF, and SL1, teristic has also been reported for Ag-NOR proteins (47). we may conclude that the rDNA transcription machinery In addition, the amount of NOR-associated proteins varies remains associated only with active NORs during mitosis. between different active NORs in human cells, and this In the inactive NORs, the rDNAs are not associated with variation correlates with NOR activity during interphase RNA pol I, UBF, or SL1. We do not know what prevents as proposed for Ag-NOR proteins (1) and for UBF (38). such associations. However, the level of methylation of the As shown for UBF (38, 53), the rDNA transcription ma- rDNAs could at least be partly responsible, since high lev- chinery is not equally distributed between active NORs. els of rDNA methylation are correlated with inactive Indeed the intensity of the labeling of positive NORs dif- rDNAs (9, 10). The presence of a repressor inhibiting such fered. Double immunofluorescence labeling experiments associations can also be proposed, but this hypothesis is showed that the different components of the rDNA tran- not presently supported by biological evidence. scription machinery vary in the same proportions in differ- ent positive NORs, suggesting a stoichiometric association of these different components. By analogy with Ag-NOR The rDNA Transcription Machinery Is Not Randomly proteins for which the amount in a given NOR is related to Distributed in Different Active NORs the number of potentially active rDNA copies in this NOR The rDNA transcription machinery remains associated (10, 48), we propose that the amount of each component The Journal of Cell BioloVgoyl,u me 133, 1996 244